Method of dividing a semiconductor wafer

Abstract

A method for dividing a semiconductor wafer which is covered by an opaque resin in a dicing process includes forming marks on the semiconductor wafer, wherein the marks are distinguished from electrodes which are formed on the semiconductor wafer. According to the method, in a dicing process, separating semiconductor chips from the semiconductor wafer can be precisely achieved.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application is a divisional application under 37 C.F.R. §1.53(b) (1) of U.S. patent application Ser. No. 09/916,320, filed on Jul. 30, 2001, now U.S. Pat. No. 6,590,274, which is a divisional application of U.S. patent application Ser. No. 09/521,888, now U.S. Pat. No. 6,303,470 filed on Mar. 9, 2000. The disclosures of this application and patent are specifically incorporated herein by reference.

FIELD OF THE INVENTION

The present invention relates to a method for manufacturing semiconductor devices, and particularly, the present invention relates to a semiconductor wafer and to method for dividing a semiconductor wafer which is covered by an opaque resin in a dicing process.

BACKGROUND OF THE INVENTION

A method has been proposed for the manufacture semiconductor devices which includes a step of covering a semiconductor wafer with a resin before a stop of separating the semiconductor devices from a semiconductor wafer.

An example is disclosed in “NIKKEI MICRODEVICES, pp. 164-167”, published on April, 1998. However, in a step of separating semiconductor chips from a semiconductor wafer which is covered by a resin according to the publication, it is difficult to precisely identify cutting points since a surface of the semiconductor wafer is covered by an opaque resin.

SUMMARY OF THE INVENTION

An object of this invention is to provide a method for manufacturing in which is possible to recognize grid lines which are hidden by an opaque resin.

To achieve the object, one aspect of the invention includes a step of forming marks on a semiconductor wafer, wherein the marks are distinguished from electrodes which are formed on the semiconductor wafer.

According to the invention, in a dicing process, separating semiconductor chips from the semiconductor wafer can be precisely achieved. That is, it is possible to realize a manufacturing process of good quality.

BRIEF DESCRIPTION OF THE DRAWINGS

While the specification concludes with claims particularly pointing out and distinctly claiming the subject matter which is regarded as the invention, it is believed that the invention, the objects and features of the invention and further objects, features and advantages thereof will be better understood from the following description taken in connection with the accompanying drawings in which:

FIG. 1 is a plane view of a semiconductor wafer according to a first preferred embodiment of the present invention.

FIG. 2(A)-FIG. 2(E) are cross sections according to the first preferred embodiment.

FIG. 3(A) is a expansion plane view of a part of the semiconductor wafer according to the first preferred embodiment.

FIG. 3(B) is a cross section at a portion of line I-I′ shown in FIG. 3(A).

FIG. 4(A) is a plane view of a first variation according to the first preferred embodiment.

FIG. 4(B) is a cross section of FIG. 4(A).

FIG. 5 is a plane view of a second variation according to the first preferred embodiment.

FIG. 6 is a plane view of a third variation according to the first preferred embodiment.

FIG. 7(A)-FIG. 7(D) are variations of a mark according to the first preferred embodiment.

FIG. 8 is a plane view of a fourth variation according to the first preferred embodiment.

FIG. 9 is a semiconductor wafer of the fourth variation which is covered by a resin.

FIG. 10 is a plane view of a fifth variation according to the first preferred embodiment.

FIG. 11 is a semiconductor wafer of the fifth variation which is covered by a resin.

FIG. 12(A) and FIG. 12(B) are cross sections according to a second preferred embodiment of the present invention.

FIG. 13(A)-FIG. 13(F) are cross sections according to a third preferred embodiment of the present invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention will be described hereinafter with reference to the accompanying drawings. The drawings used for this description typically illustrate major characteristic parts in order that the present invention will be easily understood.

FIG. 1 is a plane view of a semiconductor wafer 2 according to a preferred embodiment of the present invention. The semiconductor wafer 2 is formed with a plurality of circuit elements 1 thereon. The circuit elements 1 are semiconductor integrated circuits and are arranged in rows and columns. Grid lines 3 are arranged between the circuit elements 1. The circuit elements 1 are surrounded and sectioned by the grid lines 3. The grid lines 3 are also called scribing lines or dicing lines. A plurality of electrode pads 4 are formed on the circuit elements 1. The electrode pads 4 are electrically connected to the circuit elements 1. Also, surfaces of the electrode pads 4 are exposed in order to allow connection with conductive lines or conductive materials which are formed on the electrode pads 4.

As shown in FIG. 2(A), the semiconductor wafer 2 is covered with an insulating film 5, excepting the surfaces of the electrode pads 4. An intermediate insulating layer 6 is formed on the semiconductor wafer 2. The intermediate insulating layer 6 has openings 7 which are located above the electrode pads 4. In this preferred embodiment, the intermediate insulating layer 6 is comprised of a polyimide.

Next, a metal layer 10 is deposited on such wafer 2 by a sputtering method, as shown in FIG. 2(B). The metal layer 10 is composed of an adherent metal layer 8 and a conductive layer 9. In this embodiment, the electrode pads 4 are comprised of an aluminum or an aluminum-alloy, the adherent metal layer 8 is comprised of a titanium and the conductive layer 9 is comprised of a copper. The adherent metal layer 8 is formed for adherence between an upper layer and a lower layer. The adherent metal layer 8 can be comprised of an aluminum, an aluminum-alloy, a titanium-tungsten-alloy, or a chrome.

As shown in FIG. 2(C), a resist layer 11 is formed on the metal layer 10 and is patterned with a predetermined design. Then, a conductive line 12 is formed on the conductive layer 9 by a metallizing plating method, such as an electroplating method. The conductive line 12 is comprised of a copper and is formed in order to improve electrical characteristics and a step coverage.

Then, as shown in FIG. 2(D), the resist layer 11 is removed and a resist layer 13 is formed on the conductive line 12. The resist layer 13 is patterned and is opened at predetermined points. Bump electrodes 14 are formed in the predetermined points in the resist layer 13. The bump electrodes 14 are comprised of a copper or a gold.

Then, the resist layer 13 is removed and parts of the metal layer 10 is etched using the conductive line 12 as a mask, as shown in FIG. 2(E).

Also, as explained below, the process of the invention includes the formation of marks on the wafer 2 which are distinguishable from the bump electrodes 14.

FIG. 3(A) is a expansion plane view of a part of the semiconductor wafer 2. FIG. 3(B) is a cross section at a portion along line I-I′ shown in FIG. 3(A). The bump electrodes 14 are formed above the circuit elements 1 and marks 20 are formed on the grid lines 3. The marks 20 may distinguishable from the bump electrodes 14. In this embodiment, the marks 20 have a triangular configurations and the bump electrodes 14 have a squared configuration. However, the invention is not limited to such configurations. In FIG. 3(A), the bump electrodes 14 are regularly grouped in rows and columns and the marks 20 are out positioned outside the groups. So, even if the bump electrodes 14 and the marks 20 have the same configuration, they could be distinguished clearly.

The marks 20 can be formed simultaneously with the bump electrodes 14. If both the marks 20 and bump electrodes 14 are formed simultaneously, the resist layer 13 shown in FIG. 2(D) is also patterned and opened for the marks 20. Then, the bump electrodes 14 and the marks 20 are formed simultaneously with a metallizing plating method. In this case, the bump electrodes 14 and the marks 20 are made of the same material.

Such a semiconductor wafer 2 is then covered by a resin 21, as shown in FIG. 3(B). Top surfaces of the bump electrodes 14 and marks 20 are exposed from the surface of the resin 21. If necessary, the surface of the resin 21 is polished in order to expose the top surfaces of the bump electrodes 14 and marks 20. The surface of the resin 21 is substantially flat and is substantially level with the top surfaces of the bump electrodes 14 and the marks 20.

Then, the semiconductor wafer 2 which is covered by the resin 21 is separated along the grid lines 3 in a dicing process. As the grid lines 3 are covered by the resin 21, locations of the grid lines 3 can not be recognized directly. However, it is possible to assume the locations of the grid lines 3, since the marks 20 are exposed on the surface of the resin 21. If the semiconductor wafer 2 is divided along an extended line from the marks 20, it is possible to separate the circuit elements 1 from the semiconductor wafer 2 exactly. That is, it is possible to realize a manufacturing process of good quality.

FIG. 4(A) is a plane view of another example according to the first embodiment. FIG. 4(B) is a cross sectional of FIG. 4(A).

The marks 20 of FIG. 3(A) are arranged on the gird lines 3. However, the marks 30 can be arranged at a predetermined distance L from a center of the grid lines 3, as shown in FIG. 4(A). That is, the marks 30 are formed above the circuit elements 1.

According to the arrangement shown in FIG. 4(A), as a blade of a dicing saw which is used to separate the semiconductor wafer 2 does not touch the marks 30 directly, clogging of the blade is avoided.

Also, the marks 31 are arranged at the distance L from the center of the grid lines 3, shown in FIG. 4(A). And the marks 31 are connected to the electrode pads. That is, the marks 31 are used as one of the bump electrodes. In this case, the marks 30 have a different shape or a different size from the bump electrodes 14 in order to recognize the marks easily.

As the marks 30 can serve both as bump electrodes and marks, the step of forming separate marks can be omitted.

If the marks 31 are arranged at the distance L from the center of the grid lines 3, it is desirable to arrange a plurality of marks 40 (at least two) which are parallel with the grid lines 3, as shown in FIG. 5. As it is possible to specify the location of the grid lines 3 which are straight lines exactly, by recognizing the marks 40, it is possible to realize a manufacturing process of good quality.

Two marks can be arrange to be symmetric with respect to the grid lines 3. In FIG. 6, a pair of marks 50 are arranged at the distance L from the center 51 of the grid line 3 and are symmetric with the center 51. According to the arrangement of the pair of marks 50, it is not necessary to measure the distance L, since it is easy to specify the center of grid line using the pair of marks 50.

The marks mentioned above are shown having a triangle shape. However, other shapes can be adopted, such as a square shaped mark 55 shown in FIG. 7(A), a cross shaped mark 56 shown in FIG. 7(B), a square frame shaped mark 57 shown in FIG. 7(C) and a plurality of square shaped marks 58 shown in FIG. 7(D). These marks include segments 61 which are parallel with segments 60 of the grid lines.

According to these marks, both the location of the grid lines and the directional orientation of the grid lines can be recognized. Therefore, fewer marks are needed to recognize the grid lines, as compared with the examples mentioned previously.

The marks can be arranged at intersections of the grid lines. In FIG. 8, the cross shaped marks 56 are arranged at the intersections 70 of the grid lines 3. The marks are not limited to the cross shape, and other shapes which are described above can be applied.

FIG. 9 shows the semiconductor wafer 2 which is covered by the resin 21. Top surfaces of the bump electrodes 14 and marks 56 are exposed from the surface of the resin 21. If necessary, the surface of the resin 21 is polished in order to expose the top surfaces of the bump electrodes 14 and marks 56.

Since the resin 21 is opaque, it is difficult to recognize the grid lines 3 directly after being covered by the resin 21. However, as parts of the marks 56 are exposed in the surface of the resin 21, it is possible to indirectly identify the locations of the grid lines 3. As the marks 56 have a different configuration or size from the bump electrodes 14, the marks 56 are easy to recognize in the surface of the resin 21.

FIG. 10 shows other marks which have a different configuration and a different arrangement. In FIG. 10, four marks 81, 82, 83, 84 which are different from each other are arranged at the four corners of the circuit elements 1. The mark 81 is a squared shaped figure, the mark 82 is a triangle shaped figure, the mark 83 is a cross shaped figure and the mark 84 is an L-shaped figure.

FIG. 11 shows the semiconductor wafer 2 which is covered by the resin 21. Top surfaces of the bump electrodes 14 and marks 81, 82, 83, 84 are exposed from the surface of the resin 21. If necessary, the surface of the resin 21 is polished in order to expose the top surfaces of the bump electrodes 14 and marks 81, 82, 83, 84.

Because the resin 21 is opaque, it is difficult to recognize the grid lines 3 directly after being covered by the resin 21. However, as parts of the marks 81, 82, 83, 84 are exposed in the surface of the resin 21, it is possible to indirectly identify the locations of the grid lines 3.

According to this example, as the marks which have different configurations remain after dividing the semiconductor wafer, a direction of the circuit elements 1 can be recognized exactly. Recognizing the direction of the circuit elements is useful in specifying the plurality of the bump electrodes.

If necessary, a plurality of ball shaped electrodes are formed on the bump electrodes 14 before the dividing the semiconductor wafer.

A second preferred embodiment of the present invention is described hereinafter, referring to FIGS. 12(A) and FIG. 12(B).

A plurality of circuit elements 101 are formed on a semiconductor wafer 102. The circuit elements 101 are semiconductor integrated circuits and are arranged in rows and columns. Grid lines 103 are arranged between the circuit elements 101. The semiconductor wafer 102 is covered with a resin 104. Also, parts of a plurality of bump electrodes (not shown) are exposed in a surface of the resin 104 in order to connect conductive materials.

Since the resin 104 is opaque, it is difficult to recognize the grid lines 103 directly after being covered by the resin.

In this embodiment, locations of the grid lines 103 are recognized by an X-ray 110 which can penetrate through the resin 104 and the semiconductor wafer 102, as shown in FIG. 12(A).

Such a method is utilized for detecting differences of transmissivity of different materials. So, the materials and thicknesses of the grid lines 103 and circuit elements 101 are taken into consideration for easy recognition.

Further, as shown in FIG. 12(B), locations of the grid lines 103 may be recognized by an infrared ray or supersonic waves 111 which can penetrate through the resin 104 and can be reflected by the semiconductor wafer 102.

Also, such method is utilized for detecting differences of reflectance of different materials. So, the materials and thicknesses of the grid lines 103 and circuit elements 101 are taken into consideration for easy recognition.

According to these methods, as it is possible to recognize the grid lines which are hidden by the opaque resin, separating the circuit elements from the semiconductor wafer is precisely carried out in the dicing process. That is, it is possible to realize a manufacturing process of good quality.

A third preferred embodiment of the present invention is described hereinafter, referring to FIGS. 13(A)-13(F).

A plurality of circuit elements 201 are formed on a semiconductor wafer 202, as shown in FIG. 13(A). Grid lines 203 are arranged between the circuit elements 202.

Next, bump electrodes 204 which are connected to the circuit elements 201 and bump electrodes 205 are formed simultaneously. The bump electrodes 204 are formed on the circuit elements 201, the bump electrodes 205 are formed on the grid lines 203, as shown in FIG. 13(B). The bump electrodes 204, 205 are formed by a metallizing plating method. In this embodiment, the bump electrodes 204, 205 are comprised of copper. Also, the bump electrodes 204, 205 can be comprised of other materials, such as gold or solder.

Next, the semiconductor wafer 202 is covered by a resin 206, as shown in FIG. 13(C). Top surfaces of the bump electrodes 204, 205 are exposed from the surface of the resin 206. If necessary, the surface of the resin 206 is polished in order to expose the top surfaces of the bump electrodes 204, 205.

A resist 207 is formed so as to cover the surface of the bump electrodes 204 which are exposed from the surface of the resin 206, as shown in FIG. 13(D). The resist 207 is patterned by a general photo-etching method. An application method using a needle and a printing method can be used to form the resist 207.

Then, the bump electrodes 205 which are on the grid lines 3 are removed by a sulfuric acid or a hydrochloric acid, as shown in FIG. 13(E), in cases where the bump electrodes 205 are comprised of copper or solder. If the bump electrodes 205 are comprised of gold, a potassium iodide is used.

The resist 207 which covers the top surface of the bump electrodes 204 is removed, as shown in FIG. 13(F).

In a dicing process, as the grid lines 203 are exposed, it is easy to recognize the grid lines.

According to the third embodiment, as it is possible to recognize the grid lines, separating the circuit elements from the semiconductor wafer is exactly in the dicing process. That is, it is possible to realize a manufacturing process of good quality.

The present invention has been described with reference to illustrative embodiments, however, this description must not be considered to be confined only to the embodiments illustrated. Various modifications and changes of these illustrative embodiments and the other embodiments of the present invention will become apparent to one skilled in the art from reference to the description of the present invention. It is therefore contemplated that the appended claims will cover any such modifications or embodiments as falling within the true scope of the invention.

Claims

1. A method for dividing a semiconductor wafer, the method comprising:

providing a semiconductor wafer which has an element area containing an integrated circuit formed thereon and grid lines which surround the element area;

forming an electrode on the element area which is electrically connected to the integrated circuit;

covering a surface of the semiconductor wafer with an opaque resin;

determining a location of the grid lines by penetrating an energy through the opaque resin; and

cutting the semiconductor wafer at the determined grid lines.

2. The method for dividing a semiconductor wafer according to claim 1, wherein the determining the location of the grid lines uses a difference between materials of the semiconductor wafer and the opaque resin.

3. The method for dividing a semiconductor wafer according to claim 1, wherein the determining the location of the grid lines uses a difference between thickness of the opaque resin on the grid lines and the element area.

4. A method for dividing a semiconductor wafer, the method comprising:

providing a semiconductor wafer which has an element area containing an integrated circuit formed thereon and grid lines which surround the element area;

forming an electrode on the element area which is electrically connected to the integrated circuit;

covering a surface of the semiconductor wafer with an opaque resin;

determining a location of the grid lines by reflective infrared ray at border portions between the surface of the semiconductor wafer and the opaque resin through the opaque resin; and

cutting the semiconductor wafer at the determined grid lines.

5. The method for dividing a semiconductor wafer according to claim 4, wherein the determining the location of the grid lines uses a difference between materials of the semiconductor wafer and the opaque resin.

6. The method for dividing a semiconductor wafer according to claim 4, wherein the determining the location of the grid lines uses a difference between thickness of the opaque resin on the grid lines and the element area.

7. A method for dividing a semiconductor wafer, the method comprising:

providing a semiconductor wafer which has an element area containing an integrated circuit formed thereon and grid lines which surround the element area;

forming an electrode on the element area which is electrically connected to the integrated circuit;

covering a surface of the semiconductor wafer with an opaque resin;

determining a location of the grid lines by reflective sonography at border portions between the surface of the semiconductor wafer end the opaque resin through the opaque resin; and

cutting the semiconductor wafer at the determined grid lines.

8. The method for dividing a semiconductor wafer according to claim 7, wherein the determining the location of the grid lines uses a difference between materials of the semiconductor wafer and the opaque resin.

9. The method for dividing a semiconductor wafer according to claim 7, wherein the determining the location of the grid lines uses a difference between thickness of the opaque resin on the grid lines and the element area.